LED test / review - Fireflylite FFL707A 4000 K 95 CRI - extremely rosy, but (relatively) high performance

LED test / review EN


Fireflylite FFL707A 4000 K 95CRI


2025-02-14


  • Type: multi-die, domeless
  • Bin: unknown
  • Color group: unknown (4000 K)
  • CRI: 95
  • Rated voltage: 6 V
  • Max. Forward current: 7,200 mA
  • Max. Peak current: — mA
  • Viewing angle: — °
  • Thermal resistance: — K/W
  • Max. Temperature Tj: max. —°C

An official data sheet is not available. The above information was taken from the Fireflylite product description. No further information is available and the manufacturer is unknown.

The emitter tested here was made available to me by BLF user @thefreeman. Many thanks for this at this point.

At first glance, the FFL707A is reminiscent of an XHP70.3 HI. The biggest difference is the yellow surface, which is completely covered in phosphor. It therefore looks like a multi-die version of the FFL351A.

The four LED chips are separated by a small gap, which could have an effect on the light image.

The footprint is virtually identical to that of the XHP70.3 HI. The LED connections are wired differently using a corresponding LED board in order to achieve a forward voltage of 12 V. In this test, the FFL707A is tested in the 6 V configuration.

The four LED chips emit light to the side, with the surface of the chips having a higher color temperature than the lateral emission. The target color temperature is thus achieved by mixing the two areas.

The illuminated area is 25.8 mm2 in size. However, due to the light emitted from the sides and the gaps between the individual LED chips, it is almost impossible to determine the exact illuminated area. In this respect too, the FFL707A is pretty much the same as the XHP70.3 HI.





  • at 7,200 mA (official maximum current): 3764 lm @ 6.64 V
  • Power at official maximum: 47.8 W
  • Efficiency at 7,200 mA: 78.7 lm/W
  • Maximum reached at 15.2 A, at this point 5254 lm @ 7.63 V
  • Power at maximum 116 W
  • Efficiency at maximum 45.3 lm/W




Data for 25 °C Tsp (at 85 °C the luminance values are around 13 % lower).


Compared to the XHP70.3 HI, the performance of the FFL707A is lower. The Vf is also significantly higher. This is presumably due to the LED chips and the manufacturing quality. Nevertheless, it offers a performance commensurate with the color rendering and CCT.

The maximum is reached at around 15 A, while the XHP70.3 HI easily achieves 20 A. To achieve a good balance between efficiency and performance, the FFL707A should be supplied with a maximum current of around 9 A. Beyond this, the increase in luminous flux is marginal.

The SFT-70 does not reach the maximum current of the FFL707A by a long way, which is partly due to the smaller footprint (5050) and partly due to the thermal resistance of the LED chips.

The difference in luminance between the two LEDs more or less corresponds to the difference in luminous flux due to the virtually identical illuminated area and beam characteristics compared to the XHP70.3 HI. This LED is less suitable for longer ranges, and if it is, then only with very large reflectors. The advantage here, however, is a wide corona around the spot, which also makes it suitable for use at close range with large reflectors.



Here the beam is shown in a Sofirn IF23 Pro with OP reflector.

The gaps between the LED chips create a donut hole. Depending on the reflector, this is more or less pronounced. SMO reflectors should not be used with this LED. Otherwise there are no color fringes or other artifacts.

If an absolutely harmonious light pattern is required, a diffuser is needed in front of the reflector.



My sample tested here is incredibly rosy. The duv of -0.016 (!) is the lowest I have ever measured. According to Fireflylite, the FFL707A should have a duv of max. -0.001 ("-0.0007 to -0.001 ”). The red component is extremely high, with a peak at around 630 nm.

It should be noted that the duv can fluctuate when using optics. The spectrum was measured without secondary optics. The LED tested here was installed in the IF23 Pro as a test. In this condition, the tint corresponded approximately to that of a 219BT-V1 sw45k.

The high proportion of red distorts the colors enormously, especially in nature. Brown-red tones stand out extremely, while blue and especially green are reproduced incorrectly, especially in comparison to other high CRI LEDs.

The high proportion of red has an impact on CRI and R9. The specified CRI of 95 was massively missed and the R9 is also lower than would be expected. The CCT was maintained. With FFL LEDs you generally have to be aware that they can be very rosy in some cases and that there is above all a series variation.


  • Ra: 89
  • R9: 49
  • CCT: 4277 K
  • duv: -0.0160

The FFL707A is a very special LED and, despite the standard footprint, not for everyone. The tint (in my sample) is extremely rosy, which significantly reduces the color rendering, and the light image has a donut hole depending on the optics used, although there are no color fringes. On the other hand, it is fully compatible with the 7070 footprint and offers guaranteed color localization under BBL.

Please note that this LED is only available as a 6/12 V LED, 3 V versions are not available!


Thank you for reading this test. :slight_smile:

Greetings, Dominik


v1.0.2
21 Thanks

Thanks for the test! Interesting to see just hiw the quality of manufacturing affects the output, but you can’t really call it apples to apples against the Cree since it’s low CRI and not rosy. The light pattern though…ewww, amd this is my experience with lower quality Chineae clones of the XHP LEDs.

Good luck to find high CRI rosy 6V XHP70.3 HI :smiley:

I can only compare with LEDs I tested. Of course the XHP70.3 is much better in terms of light flux, but it still has a higher maximum current and lower Vf. This is (almost) independent from CRI rating.

3 Thanks

My numbers from Fireflly X1L Elite w/ FFL707A 4000k below. Surprised at your CRI numbers, maybe not the same LED batch/bin? My Opple LM3 is obviously hobbyist grade but in the past tested close to reviewers’ numbers wrt Ra. Lastly no donut hole with TIR.

Opple Lightmaster 3 - Turbo level:
CCT 3900k
Duv -0.0120 (even more negative at lower output)
Ra=97

Opple Lightmaster 4 - Turbo level: Ra=98, R9=91

Some other numbers: output with Texas Ace lumen tube = 2550 lm, ANSI throw 470m.

My first thought is how it highlights the flaws of CRI vs CQS or TM30. It comes as no surprise that a rosy emitter would oversaturate reds though. And lets be real here; most who go for Cree have luminous flux as their top three priorities, with tint and CRI rarely making the top five list of things they care about.

duv -0.016 is really rosy though.
Since CRI is still the most used color rendition index in the world (or at least in the flashlight community) I use this. But in any case I think, that this rosy tint is way too much. Even the rosy sw45k 219BT-V1 has not this much red in it.

Do you have taken any measurements of this LED for comparing it to my results?

2 Thanks

Indeed. My rosiest light is only around -0.013, and it borders on excessive even by my tastes.

Yes, using the same standards as everyone else makes comparison easy. Any flaws in the standard apply equally to all tests, so it cancels out. The biggest issue I have with CRI is the fact that it usually refers to Ra (R1-8), which ignores three very important colors that Re (R1-15) picks up. Last I checked, human vision was trichromatic, making red (R9), green, (R11) and blue (R12) kind of a big deal.

I don’t have any SW45K’s, but if memory serves, they’re around -0.008 or so. The only 219b light I have that low is a D3AA that puts a 2700K with a pair of 4500K’s and takes advantage of the duv drop inherent to mixed-CCTs.

Nope, but your results are close enough to my expectations of the FFL707A that I suspect we’d be within margin of error anyways.